Principal component and canonical correlation analysis of near-infrared spectra

Jie Liang; Sabine Van Huffel; Paul Casaer

doi:10.1117/12.210014

30 May 1995 Principal component and canonical correlation analysis of near-infrared spectra

Jie Liang, Sabine Van Huffel, Paul Casaer

Proceedings Volume 2389, Optical Tomography, Photon Migration, and Spectroscopy of Tissue and Model Media: Theory, Human Studies, and Instrumentation; (1995) https://doi.org/10.1117/12.210014
Event: Photonics West '95, 1995, San Jose, CA, United States

Abstract

Principal component analysis (PCA) is used to analyze the components of variable concentration (chromophores) from near IR spectroscopy measurement. Up until now, this method was only used in some special cases when the singular vectors of the measurement data matrix clearly resemble the component spectra or some simple combinations (e.g. sum or difference) of the spectra. In this paper, we find and derive a generally applicable theoretical relationship between the singular vectors and the component spectra. With this relationship, PCA becomes a powerful tool to detect how many and also which chromophores contribute to the observed attenuation changes measured by NIRS. Based on this relationship, a method is proposed to estimate the components of variable concentration from the data matrix by means of a principal component analysis and canonical correlation analysis (CCA). This method, which detects the chromophores of variable concentrations by analyzing numerically the canonical correlations, is more accurate and more generally applicable. Some results of the ananlysis are given.

Citation Download Citation

Jie Liang, Sabine Van Huffel, and Paul Casaer "Principal component and canonical correlation analysis of near-infrared spectra", Proc. SPIE 2389, Optical Tomography, Photon Migration, and Spectroscopy of Tissue and Model Media: Theory, Human Studies, and Instrumentation, (30 May 1995); https://doi.org/10.1117/12.210014

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available